ADVERTISEMENT
  About the SA Blog Network













Observations

Observations


Opinion, arguments & analyses from the editors of Scientific American
Observations HomeAboutContact

Human Brains Are Optimally Tuned for the Visual Hunt


Email   PrintPrint



eyes and brain work together to find search imageWhy is it that most of us are able to track down the tiny sketch of a bespectacled cartoon man wearing a striped shirt and a funny hat—in the midst of a busy scene filled with distractions and look-alikes?

Finding "Waldo" might be a tad tamer than hunting down a prehistoric dinner. But we can do both because we’re excellent visual hunters, and a new study demonstrates that the human brain uses its inputs and neural networks to find "search objects" at a near-optimal level.

"Surprisingly, even in a complex task like detecting an object in a scene with distracters, we find that people’s performance is near optimal," Wei Ji Ma, an assistant professor of neuroscience at Baylor College of Medicine and co-author of the new study, said in a prepared statement. "That means that the brain manages to do the best possible job given the available information"—even when "many objects might look like a target for which you are searching."

We are not unique in the animal kingdom in fixating on a target image. Other visual hunters use this strategy, too. But what is impressive is the ways in which the oversized and convoluted human brain has streamlined its abilities, the researchers contend.

"Target detection involves integrating information from multiple locations," Ma said. "It is a cognitive judgment as well as a visual one," he said—"the brain has to weigh different pieces of visual information."

For the new study, subjects were asked to quickly find a line of a particular orientation on a computer screen that displayed a whole set of helter-skelter lines with different levels of visual contrast. Subjects had less than a second to make this judgment—and the target line didn’t even appear on every screen.

"We found that even in this complex task, people came close to being optimal in detecting the target," Ma said. "The visual system is automatically and subconsciously doing complex tasks." Ma and his team suggest that in the hunt for the target line, people were using specific groups of neurons that are keyed into identifying that specific line orientation.

"Our eyes are the camera, but the process of interpreting the image in our brains is seeing," he said.

The new research was published online May 8 in Nature Neuroscience. (Scientific American is part of Nature Publishing Group.)

Image courtesy of iStockphoto/prill

Tags: ,





Rights & Permissions

Comments 9 Comments

Add Comment
  1. 1. frankblank 2:21 pm 05/8/2011

    Thanks for this. We knew.

    Link to this
  2. 2. tharriss 5:59 pm 05/8/2011

    Of course, sometimes we think we "know" something, and after studying it, find it isn’t what we thought… thus the need to occassionally study something "obvious"… just in case we’re wrong in our initial beliefs…

    Just my two cents.

    Link to this
  3. 3. zstansfi 10:01 pm 05/8/2011

    Some important questions about the research which are not addressed in this piece:

    1) The authors have defined "optimal" as the "best possible job given the available information". However, it is not clear what it is about their results that suggests humans are performing optimally. Is there a specific criterion that is being used to make this judgment?

    2) Clearly the authors varied the difficulty of the task, how did this impact performance? (Notably, this is important for determining "optimal performance" as on a very easy task we would expect performance to be at ceiling, which would preclude the determination of whether or not performance is optimal.)

    3) The authors speculate that specific groups of neurons are specialized for this sort of task. Is this based upon some of their results, or are they speculating based upon past results (e.g. single/multi-cell recordings in monkeys).

    Points #2 and #3 might be too in-depth for this sort of article, but I think point #1 is key.

    Link to this
  4. 4. rgcorrgk 1:59 am 05/9/2011

    If you are out at night driving for an extended period, at some point you will notice a car with one head light out. After this event you seem very likely to, rather quikly, see another such car. As small children traveling as pasengers, with our father at the wheel, my brother one time made this observation. He said, "you never see just one". And indeed, it seems once a visual oddity is obsevered we become sensities to it and can’t ignore it. No effort is needed, the oddity developes an unconcious attension bias.
    If you like to see old movies dont’t look for black dots at the bottom right of the screen. Once you start seeing them it’s hard not to see them – they can be down right dictracting. (PS:they were put on the old movies to add the projectioist with makeing a seemless real change.)

    R. Carlson

    Link to this
  5. 5. rgcorrgk 2:12 am 05/9/2011

    (Please discard above, to many errors, sorry)

    If you are out at night driving for an extended period, at some point you will notice a car with one head light out. After this event you seem very likely to, rather quickly, see another such car. As small children traveling as passengers, with our father at the wheel, my brother one time made this observation. He said, "You never see just one". And indeed, it seems once a visual oddity is observed we become sensitizes to it and can’t ignore it. No effort is needed; the oddity develops an unconscious attention bias.
    If you like to see old movies don’t look for black dots at the bottom right of the screen. Once you start seeing them it’s hard not to see them – they can be downright distracting. (PS: they were put on the old movies to aid the projectionist with making seamless real changes.)

    R. Carlson

    Link to this
  6. 6. iWind 7:13 pm 05/9/2011

    The black circles are in the top right of the image, and they’re on new movies too.

    From the context I have the impression that "optimal" is in comparison with other animals doing the same task with a smaller and presumably less multifunctional brain, but it’s not exactly made clear.

    Link to this
  7. 7. weiji 7:48 pm 05/9/2011

    Those are all very good questions that cannot be addressed in a popular summary, but I will do my best to answer them here.

    1) The term optimality is used in many different meanings by different researchers. We refer specifically to so-called "probabilistic optimality", which means that an observer reports "target present" when the probability that the target is present exceeds 50%. This strategy will produce the highest percentage correct. The key is thus to correctly evaluate, on every individual trial, the probability that the target is present. As a consequence, "optimal" is not the same as "perfect"; optimality is relative to the noise in the input. You can be optimal while still making mistakes, as long as you make the best possible decision given the information you are given.

    2) Yes, we varied performance by unpredictably changing the reliabilities of the stimuli. Each stimulus could have one of two reliabilities, high or low. In this way, we got many different combinations of reliabilities. The main point of this was not so much to get a range of performances (although we did), but to determine whether reliability was taken into account properly to compute the probability that the target was present (see 1).

    3) This is based on our results. We actually constructed a biologically plausible neural network that is very good at computing the probability of target presence, and is therefore very close to being probabilistically optimal. We also show that many other plausible networks do not accomplish this. We don’t make any claims about where in the brain this computation is performed, but physiologists will have opinions about that.

    If you are interested in the fine details, I would suggest to download the paper from http://neuro.bcm.edu/malab/?m=static&id=8 .

    Link to this
  8. 8. Bill Crofut 9:22 am 05/11/2011

    Re: “We are not unique in the animal kingdom in fixating on a target image….But what is impressive is the ways in which the oversized and convoluted human brain has streamlined its abilities…”

    What would be the evolutionary explanation for human uniqueness?

    Link to this
  9. 9. Wayne Williamson 7:49 pm 05/11/2011

    weiji…thanks for taking the time to reply to this article…
    based on other articles, I believe that the information the eye presents to the brain has many levels…ie movement from right to left/left to right/up/down..not just a ccd collector, it actually does alot of processing…

    Link to this

Add a Comment
You must sign in or register as a ScientificAmerican.com member to submit a comment.

More from Scientific American

Scientific American MIND iPad

Give a Gift & Get a Gift - Free!

Give a 1 year subscription as low as $14.99

Subscribe Now >>

X

Email this Article

X